This repository contains all of the source code required to reproduce our paper on ancient directional migration using an extention of smcsmc
.
pipelines/
contains two different pipelines which implement all of the analyses in the paper. The first usessmcsmc
andmsmc2
to analyse real data from the SGDP and HGDP. The second usesscrm
to simulate demographic scenarios and does inference withsmcsmc
andmsmc
. For instructions on reproducing these results, see directions in these directories. As a forewarning, these analyses are highly computationally intensive and may take several weeks or months of compute to run completely.data/
contains all of the relevant results from thepipelines/
directory used in the paper, and in the plots.r/
contains scripts for reproducing all of the plots from the paper. Note thatr/MASTER.r
, if run by itself, contains all of the code necessary to create all the plots using the data in thedata/
directory.plot/
has the output of ther/
directory, which are the plots used in the paper.tex/
has theLaTeX
source for the actual manuscript of the paper.analyses
contains configuration files for the different analyses performed using the pipelines. Similarly,lib/
has some useful files for the plotting scripts.
To reproduce the analysis completely, please install the required conda
environment. This allows us to package a majority of the software dependencies together. There are several methods to install conda
, see here for details. Once conda
is installed, create the environment for this analysis from our list of dependencies:
conda env create -f envs/aaa.yml
This will create a conda
environemnt called aaa
. Activate it:
conda activate aaa
There are a couple of dependencies that are not included in conda
and must be installed manually.
msmc2
must be installed from the Github source here.admixr
must be installed from Github as well.
install.packages("devtools")
devtools::install_github("bodkan/admixr")